Sterling trading

the intraday stop go

By Andy Webb

Jekyll and Hyde

Sterling currently exhibits
something of a split personality.
On the one hand, daily price
charts of Sterling versus most
popular currencies show a
marked tendency to trending a
characteristic that in many cases
appears to coincide (approximately)
with the introduction of the euro.
On the other, Sterlings intraday
behaviour is far less tractable,
with a tendency to do virtually
nothing for extensive periods and
then abruptly take off in a violent
directional move.
To some extent, this behaviour is
driven by the way many traders
risk profiles and time horizons
have changed over the past four

or five years. While traders might

previously have been happy to
take and hold positions in Sterling
for a month or more, this is now
far less common. This is partly
due to a lower tolerance of volatile
returns, and one way to address
this is to spend less time in the
market. As a result, many traders
now prefer to grab it and run in
order to minimise their exposure.

scramble to climb aboard the

move and benefit from what may
be the only P & L opportunity of
the trading session.

Ironically, this only serves to

exaggerate Sterlings little intraday
foibles, as an increasing number
of traders hover over their
keyboards awaiting some form of
intraday breakout. If a significant
piece of news or a substantial
flow appears, there is a frantic

Method or
emotion?
However, scrambling after the
herd doesnt guarantee that the
initial move out of the congestion
area wont prove to be a false
breakout that then develops into
another sideways trading range.
Some means of objectively
quantifying the markets

degree of congestion and

therefore the likelihood of
any breakout following through
into a substantive move is
obviously desirable.
One possible solution is the
congestion count, which counts
the number of consecutive bars
with price ranges that overlap
that of the current bar. In Figure 1
the two red horizontal lines mark

the high and low of the current

bar any consecutive preceding
bars that intersect with these
lines will be included in the
congestion count. The first,
fifth and 10th bars included
in the current congestion
count of 10 have been marked
(the congestion count always
includes the current bar).

Figure 1

There are, in fact, two versions of

the congestion count: the basic
last version (see Figure 2
bottom chart pane), which only
requires that the consecutive
preceding bars overlap the
current bar, and the group
version, which requires that
these bars must also overlap
each other in order to be
included in the count. The basic

premise is that high congestion

count values often precede major
breakouts. This certainly proves
to be the case in Figure 2, where
a last congestion count reading
of 26 on a 60 minute Cable chart
(highlighted by the vertical cursor)
is followed by a 70 tick rally. (To
give an indication of the relative
importance of a congestion
count reading of 26, the green

line overlaid on the congestion

count is its long term (1000 bar)
moving average, which shows a
reading of only 6.78).

Figure 2

False premise
However, one carefully selected
chart on its own proves nothing
taken in isolation a high
congestion count only indicates
that x number of bars are
overlapping. It is perfectly
possible to achieve a high
congestion count (particularly
with the last version of the

indicator) simply because the

last bar in a group has a wide
range. For example, the 60
minute EURGBP chart in Figure 3
shows the last version of the
congestion count suddenly
spiking to a value of 23 for just
this reason. It therefore has little
value as a trading signal, for,
by the time the congestion count
value is calculated (on the

close of the bar), the party is

already over.

Figure 3

The group version of congestion

count goes some way towards
addressing this issue. Figure 4
shows the same data as Figure
3, but with the group version of
the indicator. The requirement
that preceding bars must overlap
each other (as well as the current
one) in order to be included in
the count results in a reading of
only 4, confirming that there was

in fact no appreciable build up of

congestion. (Group congestion
count values are typically lower
than those for the last type
congestion count for example,
a group congestion count
reading above 15 in Cable would
be considered exceptional.)

Figure 4

Defining the
range 1

Figure 5 shows a pair of short entry

signals on Cable. The rules for
these signals are:

One way of ensuring that a high

congestion count is not simply
flagging a price bar with a high
range is to filter possible entry
signals by specifying the
acceptable size of the average true
range 1 of either a single bar or
series of bars.

1 The last congestion count

reading of the preceding bar
must be 20 or greater

3 The current bar crosses below

the low of the preceding bar
minus 0.05% of the close of
the preceding bar (the actual
trade entry price is shown by
the red horizontal dash
overlaid on the entry bar.)

2 The 21 period average true

range be less than 100 period
average true range

Figure 5

Although these are obviously two

rather idealised examples, the
bigger picture is also in positive
territory, albeit by not a great
deal. Figure 6 is the output table
from an evaluation of the short
entry signal outlined above,
together with its long entry
counterpart. (Rules 1 and 2 for
the long entry are the same as
for the short signal, and rule 3 is
the reverse of rule 3 for the short
entry ie. that the current bar
crosses above the high of the
preceding bar plus 0.05% of the

close of the preceding bar.) The

entry signals have been tested
on a portfolio of currency pairs
(EURGBP, GBPCHF, GBPAUD,
GBPCAD and GBPUSD) from
15 June 2002 to 20 June 2003
using 60 minute price bars.
The top row of the table shows
the total gain in percentage
terms (percentages are based
upon the entry price of each
trade 2) at the close of each bar
out as far as six bars after entry.
(For bars one to three,

performance as at the open of

each bar is also shown.) It is
important to note that these are
not complete trade system
results, but only an assessment
of the quality of the long and
short entry signals, and how
trades triggered by those signals
would perform over the bars
following entry. The lower two
thirds of the display show the
first few of the 259 individual
trade entries that are included in
the summary statistics.

Figure 6

Figure 7 shows the relative

contribution of the various
currency pairs to the overall
results. Again, the picture is

reasonably encouraging with only

one pair (EURGBP) showing a
loss over the test period.

Figure 7

Defining the
range 2
Though the basic signals have
delivered positive results (which
were confirmed by the three
preceding years back as far as
the introduction of the euro) there
is still plenty of scope for
improvement. For example, the
low average result per trade (row
two of the table) is unlikely to
cover much more than slippage
and the bid/offer spread in a live
trading situation. One of the
issues here is ensuring that the

trading range of the congestion

period is tight enough to give any
breakout sufficient impetus. (A
commonly accepted
characteristic of successful
breakouts is that they are
preceded by narrow, congested
trading ranges.)
While requiring the short term
average true range to be less
than the long term average true
range (as in the previous
example) goes some way
towards addressing this, a better
alternative might be to use the
congestion count itself as an

Figure 8

input for precisely defining the

range of the congestion period.
For example, in Figure 8 the
vertical cursor highlights a
congestion count (group
version) value of 13, which
has been used as the look back
period for the maximum/minimum
values on the chart (the maroon
and blue/green lines bounding
the price bars). The next step
would be to compare the size
of the range of the congestion
immediately prior to the
breakout with the size of previous
ranges that have the same
congestion count.

Figures 9 and 10 show the trade

signal evaluation results for such
a rule set based on the same
data as the first set of entry
signals. The additional
requirement is that the size of the
trading range on the bar prior to
entry has to be smaller than the
long-term average of all ranges
with the same congestion count
value. As can be seen, the trade

count has dropped by more than

50%, but there has also been a
dramatic improvement in
profitability for example, the
average performance of all
entries after 10 bars is +0.16%,
which for GBPUSD would equate
to approximately 27 pips at
current levels. At first glance this
may not seem particularly
scintillating, but it should be

borne in mind that these results

are just for raw entry signals
eg. there are no stop losses to
weed out poor entry signals
along the way. Figure 10 also
reveals that although GBPCAD
was by some way the strongest
performer, the other four
currency pairs also made
solid contributions.

Figure 9

10

Figure 10

11

Closing
the loop
Figure 11 takes things a step
further by providing an indication
of congestion count potential as
a tool for trading intraday Sterling
breakouts. The long and short
entry signals shown in Figure 9
have been used as the basis for
a simple trading system and
applied to GBPAUD. (The
weakest market in the entry

signal evaluation performed

above.) Only two exit signals
have been used a fixed
percentage stop loss and a fixed
percentage profit target, with
neither of the exit percentages
being optimised. The results are
based on a lot size of 1m per
trade, so the AUD185300 profit
equates to an annual return
(before slippage and execution
costs) of approximately 7.6% at
current exchange rates. Other

trade statistics are also

encouraging, such as a remove
to neutral 3 of 19.23% and
61.54% winning trades. Although
the total trade count of 26 is low,
the system also proved profitable
in the preceding years back to
the introduction of the euro (albeit
only marginally in 20012002)
and has been particularly strong
since the end of the test period
shown below.

Figure 11

12

Practical reality
While measuring the length and
range of congestion periods may
prove a useful method for coping
with Sterlings intraday vagaries,
its success is obviously
dependent upon whether there is
a tight/stable enough bid/offer
spread to actually execute trades
close to the level indicated by the

entry signals. The most

vulnerable moments in this
respect are during the transition
periods between one time zones
trading session and the next.
(Eleven of the 115 raw trading
signals shown in Figure 9
occurred during these periods.)
Fortunately the situation here is
improving rapidly, with EBS
focus hours already having a

favourable impact in the

case of EURGBP pushing the
bid/offer spread down to 2-3
pips between 07.00 and 09.00
London time. Under those
circumstances, a congestion
breakout strategy with a modest
percentage profit target
becomes viable.
February 2004

Andy Webb:
Andy Webb is a freelance writer with 20+ years' experience, most recently specialising in
derivatives, technology and trading methodologies. He has written regularly for a wide range
of journals including The Sunday Times, Sunday Business, Treasury & Risk Management,
Global Finance, Derivatives Strategy and Wall Street & Technology. He is also actively
involved in developing and programming FX trading models for several hedge funds and
proprietary trading desks.
ANDY WEBB

The views expressed in this paper are those of the author and not necessarily those of EBS.

Footnotes
1 Average true range is the moving average of true range, which is the greatest of:

High for the price bar minus the low for the price bar

High for the price bar less the close for the previous price bar

Close for the previous price bar and the low for the current price bar

2 So for example, the first individual trade signal listed (on GBPUSD) that occurred on 17 June 2002 shows a loss of 0.16%
by the close of the third bar after entry. In this case the entry price was 1.4783 (not shown) so the trade would have been
underwater by approximately 24 pips by the close of the third bar. (By contrast, the third individual entry listed GBPCAD,
dated 18 June 2002 was showing a gain of 0.64% by the close of the sixth bar on an entry price of 2.2908 = +147 pips.)
3 Percentage remove to neutral is the percentage of winning trades that would have to be excluded from a trade systems
results to reduce its total net profit to zero. A percentage remove to neutral of greater than 10 suggests that a trading
system is reasonably robust.

13